Phonological Representations in Children with Autism Spectrum DisorderRon Pomper,1 Susan Ellis Weismer,1 Jenny Saffran,1 Tristan Mahr,1 and Jan Edwards2
1 University of Wisconsin-Madison, 2 University of Maryland
• Early interventions improve language outcomes for
some children with autism spectrum disorder (ASD),1
however, 30% of children with ASD remain minimally
verbal.2 Identifying relative strengths and weaknesses in
language acquisition for children with ASD may help
tailor and improve interventions.
• Children with ASD have superior performance in non-
speech auditory tasks,3 but worse performance in
language comprehension tasks.4 Less is known about
speech perception abilities for children with ASD.5
• Do toddlers with ASD have more or less detailed phonological representations of familiar words compared to typically-developing (TD) toddlers?
Results
1 Rogers, S.J., & Vismara, L.A. (2008). J Clin Child Adolesc Psychol, 37(1), 8-38
2 Tager-Flusberg, H., & Kasari, C. (2013). Autism Research, 6(6), 479-493
3 Happe, F., & Frith, U. (2006). J Autism Dev Disord, 36(1), 5-25
4 Ellis Weismer, S., Haebig, E., Edwards, J., Saffran, J., & Venker, C.E. (2016). J Autism Dev Disord, 46(12), 3755-3769
5 Kuhl, P.K., Coffey-Corina, S., Padden, D., & Dawson, G. (2005). Dev Sci, 8(1),
F1-F12
6 Swingley, D., & Aslin, R.N. (2002). Psych Sci, 13(5), 480-484.
AcknowledgmentsThe authors have no financial or non-financial conflicts of interest. This research
was funded by the National Institutes of Health: NIDCD R01 DC012513, NICHD R37
HD037468, NICHD F31 HD091969, and NICHD P30 HD003352 core grant to the
Wiasman Center. This work was part of the Little Listeners Project:www.littlelisteners.waisman.wisc.org. Contact: [email protected]
Discussion• When collapsing across all trials, ASD and TD toddlers
were equally affected by mispronunciations, suggesting
they have the same level of detail in their phonological
representations of familiar words.
• When separating trials based on whether toddlers were
fixating the target or the distractor object at target word
onset, we find evidence that toddlers with ASD and TD
may be affected differently by mispronunciations.
• Group differences in overall word recognition accuracy
were related to verbal skills, but not nonverbal skills.
This dissociation rules out an important alternative
explanation: that toddlers who perform better on one
lab task perform better on other lab tasks (regardless of
the content).
Presented at the 2018 SRCLD Meeting
Participants
• 64 toddlers with ASD (17 female), mean age of 30.6
months (range: 24-36)
• 31 typically-developing (TD) toddlers (13 female),
mean age of 20.5 months (range: 18-24)
• Toddlers with ASD were diagnosed by an experienced
psychologist who administered ADOS-2 and ADI-R
Phonological Representations
• Assessed using a looking-
while-listening task5
• Saw pictures of two
familiar objects
• Heard a sentence
labelling one object with
either a Correct Pronunciation (CP) or
Mispronunciation (MP)
• Eye movements video recorded and coded offline
Offline Measures
• Verbal skills using Preschool Language Scales, 4th
Edition (Auditory Comprehension score) and
MacArthur-Bates Communicative Development
Inventories Words and Sentences (# of words says)
• Nonverbal skills using Mullen Scales of Early Learning
(Visual Reception scale)
References
The time courses of toddlers’ fixations to the target object were analyzed using Growth Curve Analysis (GCA).
Tests of significance were performed using model comparisons (-2*log-likelihood). The sections below analyze
the time course of fixations for all trials, then trials separated based on whether toddlers were fixating the
target object (e.g., cow) or the distractor object (e.g., shoe) at the onset of the target word (e.g., “cow”).
●●●●●●●●●
●●●●●
●●●●●●●●
●●●●●●●●●●●●●●●●●●
●●●●●●
●●●●●●●●●●●●●
●●●●●●
●●●●●
●●●●●●●●●●●●●●●●●●●●●●
●●●
●
●
●
●●
●
●●●●●
●●●●
●●
●●●●
●●●●●●●●
●●●●●●
●●●●●
●●●
●●●●●●●●●●
●●●●●●●●
●●●●●●
●●●●
●●●●●●●
●●●●●●●●●●●
ASD TD
300 600 900 1200 1500 1800 300 600 900 1200 1500 1800
−1.00
−0.75
−0.50
−0.25
0.00
0.25
0.50
0.75
1.00
1.25
1.50
1.75
2.00
Time since target word onset (in ms)
Fixa
tion
Empi
rical
Log
it
●
●
CPMP
lmer(elog ~ (ot1+ot2+ot3)*Condition*Group + ((ot1+ot2+ot3)*Condition|Sub.Num), data=d.gca, weights=1/wts, control=lmerControl(optimizer='bobyqa'),REML=FALSE)
Toddlers were less accurate in fixating the target object when it was labelled with a MP (e.g., “gow”) compared to a CP (e.g., “cow”).• Significant effect of Condition on t0, t2, and t3,
!2(1)’s > 5.6, p’s < .02
Toddlers in the ASD Group were less accurate in fixating the target compared to the TD Group.• Significant effect of Group on t0, t1, and t2,
!2(1)’s > 10.4, p’s < .002
The effect of Mispronunciations on accuracy was the same for toddlers in both Groups.• Non-significant effect of Condition:Group on t0, t1, t2, and t3,
!2(1)’s < 2, p’s > .16
● ● ●●
●
●
●
●●
●●
● ●
●●
●●
●●
● ● ● ● ● ● ● ●● ● ● ● ●
● ● ●● ●
● ● ●●
●
●
●
●
●
●●
●●
●●
●● ● ● ●
● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●
● ● ● ●
●●
●
●
● ● ●● ● ●
● ●● ● ●
●●
●●
●●
● ● ● ● ● ● ●●
●● ● ●
● ● ●
●
●
●
●
●
●
●
●●
●
● ●● ●
● ● ● ● ● ●● ●
●●
●● ●
● ●● ● ●
● ●
ASD TD
300 600 900 1200 1500 300 600 900 1200 1500
−1.0
−0.5
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
Time since target word onset (in ms)
Fixa
tion
Empi
rical
Log
it
●
●
CPMP
● ●●
●
●
●
●
●●
●
●
●●
●
●●
● ●●
● ●●
●●
●●
● ● ●● ● ● ● ● ● ● ●
● ● ●●
●●
●
●
●
●●
●●
●●
●●
● ●● ●
● ●●
●● ●
●● ● ● ● ● ● ● ● ●
● ●●
●
●
●
●
●
●
●
●
●
●
●
●●
●●
●
●
●●
●●
●● ● ●
● ● ● ● ● ● ● ● ●
● ●●
●
●
●
●
●
●
●
●
●
●
●●
●●
●
● ●●
●● ●
●●
● ● ● ●● ● ● ● ● ● ●
ASD TD
300 600 900 1200 1500 300 600 900 1200 1500
−3.5
−3.0
−2.5
−2.0
−1.5
−1.0
−0.5
0.0
0.5
1.0
1.5
2.0
Time since target word onset (in ms)
Fixa
tion
Empi
rical
Log
it
●
●
CPMP
All Tria
lsTa
rge
t-Initia
l Trials
Distra
ctor-In
itial Tria
ls
For target-initial trials, there was a strongereffect of Mispronunciations for toddlers in the TD Group compared to ASD Group• Significant effect of Condition:Group on t0 and t2,
!2(1)’s > 8.6, p’s < .01
No effect of Mispronunciations for toddlers in the ASD Group• Non-significant effect of Condition on t0 and t2,
!2(1)’s < 2.3, p’s > .13
…but a significant effect for toddlers in the TD Group• Significant effect of Condition on t0 and t2,
!2(1)’s > 19.9, p’s < .001
For distractor-initial trials, the effect of Mispronunciations was the same for toddlers in both Groups• Non-significant effect of Condition:Group on t0, t1, t2, and t3,
!2(1)’s < 1.4, p’s > .24
Significant effect of Mispronunciations for toddlers in the ASD Group• Significant effect of Condition on t0 and t2,
!2(1)’s > 7.4, p’s < .01
…and in the TD Group• Significant effect of Condition on t0,
!2(1) = 7.4, p < .01
After controlling for differences in verbal skills:• toddlers in the ASD Group were equally accurate in
fixating the target object compared to the TD Group• Non-significant effect of Group on t0, t1, t2, and t3,!2(1)’s < 1.1, p’s > .29
• the effect of Mispronunciations was the same for
toddlers in both Groups• Non-significant effect of Condition:Group on t0, t1, t2, and t3,!2(1)’s < 1.8,
p’s > .18
After controlling for differences in nonverbal skills:• toddlers in the ASD Group were less accurate in fixating
the target object compared to the TD Group• Significant effect of Group on t0, t1, and t2,!2(1)’s > 4.8, p’s < .05
• the effect of Mispronunciations was the same for
toddlers in both Groups• Non-significant effect of Group on t0, t1, t2, and t3,!2(1)’s < 1.3, p’s > .25
Offline MeasuresIntroduction
Method
Top Related